Web applications have become important business, entertainment and social tools. Commerce is conducted over the Internet via merchandise transactions, financial transactions, etc. Internet communication typically involves a client computer system, e.g., having a browser, communicating with a remote server system. The server system typically has installed thereon a web application program. The client system transmits page requests to the server system which responds with page responses to create a web experience for the client user. The requests and responses are called “network traffic.”
Designers of web application programs would like to gain information about client experience including the client behavior and application behavior when interacting with a web application in order to better serve the client. In order to understand the manner in which users interact with a web application, prior art systems have been developed that are able to passively monitor and store the network traffic (without disturbing the web application program or the client) and further these passive systems can separate and store the data by sessions that are unique to a particular client. By analyzing the recorded network traffic, important client behavior may be understood. Also, by recording the network traffic by sessions, a specific client interaction can be replayed to an operator during a client telephone call in order to troubleshoot a web application program or determine how a client can be better assisted.
Unfortunately, it is not an easy task to modify a web application program to take advantage of information that may be obtained by passive monitoring and recording of the network traffic. Server based systems can be very complex and large. Many companies that run such servers resist frequent modification of them and their resident web application programs.